TOWN LINE BROOK URBAN WATERSHED STUDY MODELING INCREMENTAL IMPROVEMENTS

Transcription

1 TOWN LINE BROOK URBAN WATERSHED STUDY MODELING INCREMENTAL IMPROVEMENTS Marcus M. Quigley, P.E., Project Engineer, GeoSyntec Consultants Steven P. Roy, Associate, GeoSyntec Consultants Lawrence Gil, North Coastal Watershed Team Leader, Massachusetts Executive Office of Environmental Affairs GeoSyntec Consultants, Inc. 629 Massachusetts Ave Boxborough, MA ABSTRACT Innovative approaches that are available for addressing stormwater pollution and flooding problems in highly urbanized areas are often not proposed as alternative solutions due to the complexity of analyzing the marginal benefit of a large number of low cost alternatives in favor of robust and often costly engineering solutions. Town Line Brook which drains significant potions of the cities of Revere, Malden, Everett, and Melrose just north of Boston, Massachusetts has repeatedly been the victim of oversimplified (and costly) proposed solutions to a complex series of water quality and quantity problems. This paper discusses the recent and ongoing efforts of the authors to address these problems in this 2.5 mile long tidal creek draining approximately 2500 acres of highly urbanized area. A proposed range of innovative approaches are proposed including: restoration of floodplain function through the creation of offline storage, salt marsh and freshwater wetland rehabilitation, self-regulating and conventional tide gate installation and optimization, in-channel sediment removal, bank and channel stabilization, erosion control, and removal and rehabilitation of engineered structures. Watershed hydrology and hydraulics have been modeled using a continuous simulation (based on 50 years of historical hourly tide and rainfall data) of both the main channel and the complex drainage system utilizing the SWMM model coupled to a project GIS. The methods used have significant implications for similar locations nationally demonstrating that the difference between inaction and implementation can lie in our willingness to embrace innovative and incremental solutions to complex water quality and flooding problems. INTRODUCTION The Massachusetts Environment Trust (MET) established this study of the Town Line Brook watershed and its two tributaries (Trifone Brook and Linden Brook) to provide recommendations for improvements to the existing drainage infrastructure and management of the watershed in order to address public safety hazards created by chronic flooding of the brooks and reduce pollution entering the Pines River and surrounding shellfish beds. 78

2 This paper focuses on the technical approach used in the study for assessment of low cost/high return flooding mitigation, channel and wetland restoration, and water quality improvement alternatives. PROJECT SITE AND HISTORY The Town Line Brook watershed is comprised of approximately 2500 acres of highly urbanized areas of four towns (Revere, Malden, Everett, and Melrose) located just north of Boston on the coast of Massachusetts (see Figure 1). Population densities in these three towns range from 16.5 per acre to 9.3 per acre. Linden Brook Rumney Marsh Route 1 Tide Gates Watershed Boundry Town Line Brook Figure 1. Site map showing watershed, main drainage channels, receiving water for the Town Line Brook watershed. The main channel of Town Line brook is approximately 2.5 miles long. Until the late 1950s, the channel was under tidal influence. As part of a flood mitigation project that was intended to include a sizable detention facility and pump station, the upper reaches of the channel were excavated and concrete lined. At the same time, the major tributary of Town Line Brook, Linden Brook, which drains 1100 acres of the watershed, was almost completely enclosed in a system of culverts. The proposed detention facility and pump station were never constructed. However, tide gates were placed at the most downstream culvert to limit tidal flows into Town Line Brook and back into the drainage system. 79

3 Under current conditions, the main channel drains through a set of tide gates to Rumney Marsh, a state designated Area of Critical Environmental Concern (ACEC). The main channel and the tributary drainage system are subject to partial tidal influence as a result of the installation of a series of self-regulating and conventional tide gates (SRTs). The rehabilitated tide gate system was constructed in 2001 for the restoration of upstream salt marsh areas and flood protection. The tide gate structures present prior to 2001 were of a conventional design and in poor operating condition, which resulted in minimal flood protection and regular inundation of historical salt marsh areas. Watershed hydrology has changed dramatically over the past 70 years due to the extensive development that has taken place. In addition, the floodplain has rapidly disappeared in this century. Field observations of encroachment activities are supported by historical orthophotographs of the area, which were obtained by the authors for each of the past seven decades. Tidal fluctuations in the receiving water vary greatly with the astronomical tidal fluctuations being in excess of 12 feet and storm surges in excess of 14 feet (above mean lower low water) during extreme events. METHODS Model Selection Based on the complexity of the hydrologic and hydraulic situation in the watershed (downstream tidal boundary conditions with self-regulating tide gates and upstream urban runoff) and the desire to understand the impacts that a variety of flood mitigation approaches would have on the frequency of flooding, the SWMM model was selected for modeling hydrology and drainage system and channel hydraulics. The authors felt that in order to provide stakeholders with some tangible evaluation of mitigated impacts under the conditions present in Town Line Brook and Linden Brook, a continuous simulation model would need to be developed. The continuous simulation approach allowed the authors to provide frequency analysis results on actual water surface elevations and flow rates in the Brook and drainage system as well as elevation-duration curves. In addition, the use of the SWMM model allowed the authors to examine the frequency and duration of frequent events to a much greater extent than would have been possible with steady state or single event models. Where tidal boundary conditions drive flooding, it is often difficult to assess the impacts of proposed flood mitigation alternatives for frequent events without a continuous simulation model. The typical modeling approach is to determine water surface profiles resulting from a synthetic storm event occurring during a specific tidal condition (e.g., flows resulting from the 50-year 24-hour rainfall combined with 50-year tidal elevations). Use of the SWMM model also allows for frequency analysis of events that result in the majority of pollutant loads to the receiving waters (i.e., smaller events that are often not of specific interest for flood mitigation.) 80

4 Model Setup The SWMM model was developed from a variety of existing sources of information including: Historical hourly rainfall records (50 years) Historical hourly and 6 minute tidal data (80 years) GIS based land use data GIS based digital (or digitized hard copy) soil survey maps (MassGIS and National Resources Conservation Service) Digital elevation models and other digital topographical data of the watershed (United States Geological Survey, Massachusetts Highway Department, and MassGIS) Existing HEC-2 cross-sectional data for the main channel (previous study by Haden and Wegman) Scanned and digitized drainage maps (from Towns of Revere, Malden, Everett, and Melrose) Original design plans for the culverts and main channel improvements (obtained from the Metropolitan District Commission) Digital Orthophotography (MassGIS) Field survey, measurement, and documentation of channel sections, inverts, and outfalls. Hydrology for the model was carried out using historic rainfall records in the SWMM Rain and Runoff modules. The Runoff module was developed based on available land use data, digitized soils data, digital elevation information, and other pertinent available information from a variety of references and sources (e.g., impervious percentages for land use categories as developed by MassGIS). The Runoff module contains 70 sub-watersheds. Hydraulics were simulated by employing the Extended Transport (Extran) Module. Open channel hydraulics for the main channel were simulated utilizing irregular cross-section data available from existing HEC-2 models and new sections developed based on one-foot contour interval digital topographical maps. Pipes less than 24 in diameter were excluded from the SWMM model in order to strike a balance between accurately representing the drainage system and model complexity. The Extran module contains 125 conduits including open channel sections. Water Quality Issues Although Town Line Brook is subject to the typical suite of omnipresent urban pollutants, the primary pollutant of concern in Town Line Brook is fecal coliform contamination due to the large areas of shellfish beds downstream in the receiving water (Rumney Marsh). Water quality improvement opportunities for pollutants resulting from non-point sources in the watershed and Town Line Brook itself are limited by a number of factors, including availability of land. Typical approaches such as providing regional facilities for the removal of pollutants are limited in their applicability. Dry weather sources are being actively pursued in the watershed. Although illicit connections and leaking sanitary sewers may play a major role in the current water quality impairment, addressing these sources alone will most likely not bring the water quality of Town Line Brook below required levels (200 mpn/100ml) due to the abundance of 81

5 non-point sources. The authors are currently conducting a wet weather and dry weather monitoring study to help identify wet weather sources and design both structural (constructed wetlands) and non-structural (public education and pet waste management programs) Best Management Practices (BMPs) aimed at reducing discharges of non-point fecal coliform pollution. The work conducted as part of the hydrology and hydraulics study are useful for better understanding the contributions of source areas and impacts on non-point source loads from proposed structural BMPs. Assessment of Flood Mitigation Alternatives One of the primary objectives of the project was to identify strategies and modifications to the drainage system that would result in decreased flooding (frequency and duration) in Town Line Brook. Mitigation strategies broadly fall into two categories in the alternatives analysis according to their assessment methodology: 1) strategies that can be quantitatively evaluated in the SWMM model; and 2) strategies that can be evaluated qualitatively or through non-modeling approaches. Alternatives considered both separately and in combination include: Install additional conventional tide gates at a variety of locations in the drainage system (modeled). Provide new main-stem offline storage and improved wetland and salt marsh environments (modeled). Diversion of flows from Linden Brook to existing wetlands for water quality improvement or complete diversion for flood control (modeled). Use of portions of the ACEC for water quality facilities and flood storage (modeled). Adjustment of the SRT closing setting and active management of the tide gates during and before extreme events. Channel dredging (capacity analysis) Channel removal and rehabilitation (qualitative evaluation). Upstream storage (qualitative evaluation and limited analysis). Development and Zoning (qualitative analysis). Increasing the height of dikes to protect the floodplain from large storm surge. In addition to quantitative and qualitative assessment of the above alternatives, options were evaluated for their ability to be implemented. Some options were excluded prior to the alternatives analysis due to extreme cost and the historical precedent of the failure of proposed large-scale solutions. As a guiding principal, alternatives that had potential to increase flooding in any part of the watershed above current conditions were excluded. Modeling of specific historical extreme events under alternative strategies identified a number of options that failed due to this criterion. Specifically, placement of tide gates at the Squire Road culvert (See Figure 2) would provide increased protection in some sections of the watershed (upstream of the Squire Road culvert), but also increased flooding in the Linden Brook Culvert. 82

6 Figure 2. Map of Lower Reaches of Town Line Brook and Linden Brook Although this alternative would prevent the storage of tidal flows in the channel above Squire Road, providing additional runoff storage volume, the increased capacity of the system is not felt by the Linden Brook watershed; in fact the flood storage volume of the channel and floodplain available to Linden Brook flows is significantly decreased. This is a result of the relative size of the area draining above and below the proposed tide gate location and the available storage capacity in the two sections of the channel. Placement of tide gates at Squire Road prevents storage of runoff from the 1100 acre Linden Brook watershed in areas above the potential tide gate location. Results such as these are difficult to demonstrate in this watershed without the use of the continuous simulation model. In fact a number of alternatives increased flooding during historical extreme events. A summary of some of the results from modeling a variety of flood mitigation alternatives under extreme historical conditions are provided in Table 1. 83

7 Table 1. Modeling Results for Flood Mitigation Alternatives During Historical Extreme Events February (2.86 inches 100-year tide) Main Above Tidal Channel Linden Squire Elevation Above Rt. 1 Culvert Road June (6.77 inches No surge) Above Squire Road Tidal Main Channel Linden Elevation Above Rt. 1 Culvert Scenario Maximum Water Surface Elevation (ft, NGVD) Maximum Water Surface Elevation (ft, NGVD) Current Conditions Lower SRT Setting Elev Closed SRT Conventional Tide Gates at Squire Road Tide Gates at Squire and Offline Storage Upstream of Squire Road Storage Above Squire Road Only Complete Diversion of Linden Brook, Additional ACEC Wetland Storage for Linden Watershed (additional 4.6 ac), Tide Gates at Linden, Squire Road Tide Gates, Offline Storage Above Squire Road RESULTS AND DISCUSSION The intent of this project was to identify flood and water quality improvements that could be implemented for Town Line Brook. The basic premise of the project was that innovative solutions that provided significant marginal benefit, but may not solve all of the problems in Town Line Brook could be found through continuous simulation modeling of hydrology and hydraulics. The authors found through modeling and qualitative analysis that several such solutions that could be implemented in combination to provide a noticeable improvement in not only flooding, but also water quality, and habitat. These solutions were compiled into a preferred approach. The preferred approach consists of the following: Install tide gates at the Linden Brook culvert to make available additional storage (as much as 10 to 13 ac-ft) at high tide when the SRTs are not set closed. Install tide gates on Trifone Brook culvert to protect upstream areas from excessive downstream water surface elevations. Set SRTs to close at elevation 2 NGVD (they are currently permitted to close at 4 during the winter months and 5 during the summer). Create approximately 60 ac-ft of offline storage on the main channel in combination with wetland restoration consistent with adjusted SRT closing elevation. Dredge the channel of approximately 4000 cubic yards of sediment that have accumulated in lined reaches. Increase flood dike height to 9 NGVD at all locations. A number of the alternatives can be implemented independently. Specifically, installation of conventional tide gates at the Linden Brook culvert and at Trifone Brook can be carried out 84

8 independently of other components. The most notable improvement in flood elevations results from the combination of setting the SRTs to low closing elevations and providing offline storage. SRT closing elevations cannot be lowered until significant upstream wetland rehabilitation is conducted to account for major changes in water elevations that result from the lower setting. In addition, the construction of offline storage areas should be conducted as an integral component of the wetlands/salt marsh restoration. The construction of offline storage areas also could include channel modifications such as removing sections of the channel to allow flows to enter into offline areas. The goal for the channel modifications would be to restore a more natural flow regime with interactions between the main channel and adjacent salt marsh offline storage areas. It is expected that if properly designed, offline storage areas and the wetland restoration work could aid significantly with water quality. The SWMM model demonstrates the impacts of using the preferred approach. An elevationduration curve provides a useful overview of the effects of the strategy (see Figure 3) on water surface elevations. Residential flooding occurs above elevation 6. Figure 3 demonstrates that during the 10 year modeled period ( ) the maximum water surface elevation upstream of the SRTs in the main channel is decreased from 6.5 feet under current conditions to 5.2 feet under the preferred alternative. It is also important to note that a considerable level of protection has been reached through the installation of the system of tide gates installed in Water surface elevations would be much closer to the downstream water surface elevations in Rumney Marsh if the tide gates had not been rehabilitated. 100% Maximum Mitigated 5.2 ft Maximum Current Conditions 6.5ft Maximum Tide 8.9ft 90% 80% Pe rc en t of To tal Du rat io Percent of Total 70% 60% 50% 40% 30% Water Surface Immediately Upstream of Route Current Conditions Downstream Water Elevation ( Rumney Marsh) Water Surface Immediately Upstream of Route Mitigated 20% 10% 0% Water Surface Elevation (ft, Figure 3. Percent of Total Model Duration where Water Surface Elevation is at or Below Indicated Level During Model Run During 11 Year Period through

9 ACKNOWLEDGEMENTS The authors would like to thank the large number of active project participants at the following agencies and organizations: Massachusetts Environmental Trust, Saugus River Watershed Council, the Towns of Malden, Revere, Everett, and Melrose, Massachusetts Department of Environmental Protection, Environmental Protection Agency (Region 1), Massachusetts Highway Department, Massachusetts Coastal Zone Management, and the Massachusetts Executive Office of Environmental Affairs. The success of this project is directly linked to the willingness of the MET oversight committee to embrace innovative approaches and seek solutions that provide high marginal benefit. This project was funded by the Massachusetts Environmental Trust. 86